Introduction

Once accepted as a legitimate judicial practice, torture has come to be widely condemned as unacceptable. The atrocities of World War II led the framers of the 1948 Universal Declaration of Human Rights to include a prohibition against torture, stipulating in unqualified terms that “no one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment” (Article 5). Similarly, the Geneva Conventions, which were expanded and revised in 1949, not only provided protection for prisoners of war and civilians but also banned the use of torture and cruelty against “unlawful” combatants as “outrages against personal dignity” (Fourth Geneva Convention, Article 3). Since that time, various international conventions have made the ban on torture an absolute moral imperative, assigning it the status of a peremptory norm (jus cogens) that is widely considered to be binding on all states, whether they have ratified a particular treaty or not. The 1966 International Covenant on Civil and Political Rights prohibited torture even “during public emergencies that threaten the life of the nation” (Articles 4 and 7). Similarly, the 1984 Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment insisted that “no exceptional circumstances whatsoever, whether a state of war or a threat of war, internal political instability or any other public emergency, may be invoked as a justification of torture” (Article 2). In 1998 the International Criminal Tribunal for Yugoslavia ruled in the case of the Prosecutor v. Anto Furundzija that the jus cogens value of the prohibition against torture meant that national measures authorizing or condoning torture or absolving perpetrators through amnesty laws are impermissible; furthermore, the court ruled that every state is entitled to investigate, prosecute, and punish or extradite individuals accused of torture who are present in a territory under its jurisdiction (de Wet, 2004, p. 98). This ruling was upheld in the landmark Pinochet case, in which the British House of Lords divested former Chilean dictator Augusto Pinochet of his sovereign immunity and ruled that even heads of state can be held accountable for violating crimes against humanity, including the prohibition against torture (Evans, 2006).

Yet just as considerations of political expediency led the British government to allow General Pinochet to return to Chile rather than be extradited to Spain to stand trial, governments and international courts have not consistently upheld the prohibition against torture. Moreover, torture continues to be practiced by many countries throughout the world, including leading democracies. In 2014, 155 countries have ratified the Convention Against Torture, but instances of torture or other ill-treatment were documented in 141 countries (Amnesty International, 2014). Few countries openly acknowledge employing such practices, resorting to a variety of strategies to circumvent the legal prohibition against torture, including denials that given treatments constitute torture, plausible deniability, and torture by proxy. For example, though the George W. Bush administration denied using torture against detainees in the “war on terror” launched after September 11, 2001, its clandestine detention program and use of brutal interrogation techniques defied such public claims (Senate Select Committee on Intelligence, 2014; for a fuller discussion, see Evans 2019).

International law and universal human rights norms symbolize modern progress toward banning the use of torture as well as cruel, inhuman or degrading treatment or punishment. Yet the continued use of torture suggests that protections remain vulnerable to state interests. Although torture is no longer used as a spectacle of state power, it continues as a practice hidden in plain sight. By downplaying the physical and psychological harm caused by interrogation methods and questioning the credibility of victims, governments maintain legal and moral respectability while simultaneously inflicting tortuous experiences in an attempt to uncover information to prevent future attacks.

This article begins by laying out the legal definition of torture and a number of complexities and ambiguities that arise from this definition. Next, it turns to a discussion of the history of torture, analyzing its shift from an open and accepted practice to one that became hidden and stigmatized. The article then summarizes recent scholarship on institutional and legal influences on the practice of torture, pointing to limits on the effectiveness of democratic and judicial institutions in restricting torture. It concludes with reflections on the social forces that enable torture to continue. Although people abhor the general idea of torture, they are willing to believe official denials that authorities actually engage in torture, especially when this is hidden in prisons or black sites overseas. Moral disengagement and victim blaming make it easier to condone violence, while social amnesia contributes to a lack of accountability that suggests that abuse is acceptable.

The Definition of Torture

The UN CAT (Article 1) defines torture as:

any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity. It does not include pain or suffering arising only from, inherent in, or incidental to lawful sanctions.

According to international law, torture is the intentional infliction of intense suffering aimed at forcing someone to reveal information, punishing unwanted behavior or inspiring fear in a broader population. While this legal definition focuses on the involvement of states and their agents rather than cruelty committed by private citizens, scholars have noted that illegal organizations, such as the mafia or guerrilla armies, are also capable of torture (Davis, 2005, p. 163).

Since freedom from torture is defined as a universal human right which states may not violate under any circumstances, states go to great lengths to insist that their conduct does not qualify as torture. As legal scholar John Parry (2005, p. 520) comments, “governments will interpret the legal definition of torture to permit specific forms of coercion because they are ‘not torture,’ a move which allows formal adherence to an absolute ban on a shrinking category of conduct.” Governments have carefully parsed the language with which they describe interrogation techniques, using euphemisms and narrow legal interpretations to differentiate between their methods and torture. Governments acknowledge employing “enhanced interrogation techniques” but insist that these do not constitute torture. Like many legal formulas, the definition of torture is sufficiently vague as to permit governments to draw semantic distinctions; governments can thus deny that the harsh interrogation methods that they employ do not meet the high threshold that legally defines torture. To qualify as torture, interrogation methods must not only cause pain but “severe” pain; they must not only involve degrading and inhuman treatment, but an “aggravated” form of such treatment. The prohibition against the intentional infliction of severe physical or mental suffering therefore opens up a subjective judgment as to whether particular methods cause sufficiently severe pain as to be considered torture. As Tobias Kelly (2012, p. 170) writes, “for an act to be considered torture . . . the intensity of pain and suffering is of central importance. Yet the law provides no precise point at which pain tips over into severe pain and an act becomes torture.” As a result, even though the prohibition against torture is absolute and definitive in principle, it is quite difficult in actual practice to recognize and document when torture has taken place (Kelly, 2012, p. 169). This is especially true given the legal reservations that countries like the United States have adopted when ratifying the UN CAT, effectively excluding practices like sensory disorientation and self-inflicted pain and requiring specific intentionality that is “situated subjectively in the mind of the torturer, not objectively in the nature of his actions” (McCoy, 2011, p.34). Though the use of excessive force would seem to be banned under international law and the US Constitution, strict textual analysis suggests that exceptions are permitted if the use of force is deemed to be “reasonable,” is applied as part of a “good faith effort to restore or maintain discipline,” or is justified by “compelling government interest” (Parry, 2005, pp. 527–528).

Public authorities have a vested interest in insisting that instances of painful abuse should not be labeled as torture, given the political and legal consequences this carries. Following the attacks of September 11, 2001, legal memoranda prepared by the Office of Legal Counsel within the Department of Justice authorized the use of harsh interrogation practices against detainees in Afghanistan, Iraq, and elsewhere, maintaining that these did not technically constitute torture. According to an August 2002 memo, abuse does not rise to the level of torture unless it inflicts pain “equivalent in intensity to the pain accompanying serious physical injury, such as organ failure, impairment of bodily function, or even death.” For purely mental pain or suffering to constitute torture, it had to result from “extreme acts” and “result in significant psychological harm of significant duration, e.g., lasting for months or even years.” The memo went on to examine a number of techniques such as sensory deprivation that “may amount to cruel, inhuman or degrading treatment,” but insisted that these “do not produce the pain or suffering of the necessary intensity to meet the definition of torture” (“Standards of Conduct for Interrogation under 18 USC §§2340–2340A,” August 1, 2001; reprinted in Greenberg and Dratel, 2005, pp. 172–217). Public opinion in the United States suggests that most Americans generally agreed with such distinctions; for example, after the shocking pictures of prisoner abuse at Abu Ghraib prison in Iraq were shown in April 2004, large majorities opposed the kinds of techniques used in Iraq but only a third agreed that what US soldiers did to prisoners there amounted to torture (Morin & Deane, 2004).

Such distinctions allow officials to distance themselves from any association with torture. In the popular imagination, torture calls to mind images of dungeons furnished with medieval instruments designed to maim and mutilate. As such, torture tends to be associated with barbaric methods of inflicting physical pain that are readily condemned by the civilized world. Yet modern torture, as Darius Rejali (2007, pp. 1–5) chronicles, often takes the form of “clean” torture techniques that inflict pain without leaving visible marks. Techniques such as electric shock, choking with water, near asphyxiation, and certain types of beatings are specifically designed not to “bruise the merchandise,” which makes them easier to deny and allows states to avoid bad publicity (Rejali, 2007, pp. 1–5). Interrogation techniques euphemistically referred to as “sleep management” (20-hour interrogations for every 24-hour cycle), “environmental manipulation” (exposure to extreme heat or cold), “stress and duress” (forcing a prisoner to stand or sit in uncomfortable positions for long periods of time), and “ego down” (degrading treatment) are often downplayed as “torture lite,” suggesting that they are not as harmful or severe as “real” torture (Bowden, 2003, p. 53; Jaffer & Singh, 2007, pp. 8–19; Wolfendale, 2009, p. 54). The use of “acoustical bombardment”—playing extremely loud, dissonant music or white noise over long periods of time—has been used as a means of sensory deprivation, preventing prisoners from being able to think straight and compounding their disorientation through sleep deprivation. Although the use of music in interrogations may at first seemed innocuous, it has been found to have debilitating effects on prisoners (Cusick, 2013, pp. 275–291). Moreover, the use of songs—like the theme song from Sesame Street or Barney the Purple Dinosaur—seems to “undercut the seriousness of the effects of the physical and psychological abuse of which it is a part. We’re able to laugh off the form of abuse on account of the associations of its content. And once we laugh at it, we effectively, although almost surreptitiously, break through the absolute ban on torture” (Cusick & Joseph, 2011, pp. 17–18).

The various techniques used as part of enhanced interrogations are not activities that one would intuitively think of as being torturous and may even be seen as normal. For example, Rudy Giuliani, the former mayor of New York City, compared sleep deprivation to the “fatigue of campaigning” that he experienced (Bellaby, 2015, p. 185). When asked to approve specific interrogation techniques for prisoners at Guantánamo, Defense Secretary Rumsfeld authorized forcing prisoners to stand for long periods of time and asked why this was limited to eight hours per day, suggesting that he stood that long if not longer himself (Haynes, 2002). Such responses suggest that one knows what it feels like to undergo one of the torture lite techniques and that detainees are weak for not being able to withstand them (Cusick & Joseph, 2011, p. 18). However, as Elaine Scarry (1985, pp. 47–48) points out, forcing people to stand completely immobile for long periods of time “can produce as violent muscle and spine pain as can injury from elaborate equipment and apparatus, though any of us outside this situation, used to adjusting our body positions every few moments before even mild discomfort is felt, may not immediately recognize this.” Ross Bellaby (2015, pp. 185–186) notes that interrogation techniques are used in combination and over long periods “with no option of recovery, revival or understanding of when the ordeal will end. . . . Over time the building up or continuation of such activities can have profound effects on an individual’s mental, physical and emotional state” and in combination amount to full torture.

Nonetheless, by employing “harsh interrogation” rather than torture, interrogators can more readily see themselves as “professionals motivated by the need to gain intelligence essential for saving lives” rather than as brutal or sadistic torturers (Wolfendale, 2009, p. 49). Moreover, the use of “lite” techniques is easier to justify because interrogators are not forced to inflict physical violence directly. While they may force detainees into stress positions or subject them to sensory deprivation, the agony that the detainees experience stems from their prolonged exposure to certain conditions rather than an immediate response to the infliction of pain (Wolfendale, 2009, p. 55). Thus, it is easier to convince interrogators that their actions are benign and that the detainees are responsible for their own pain (Wolfendale, 2009, pp. 56–57).

Thus, according to certain perspectives, interrogation techniques and detention procedures that aim at “softening up” prisoners through sensory deprivation, sexual humiliation, and exploitation of phobias, though unpleasant, do not constitute torture. Similarly, psychological manipulations designed to break a person’s resistance by creating fear, terror, or helplessness may not be counted as torture because they are presumed not to cause prolonged mental harm, such as the development of post-traumatic stress disorder (Başoğlu, Livanou, & Crnobarić, 2007, pp. 277–278). Such a distinction between physical torture and extreme mental suffering is reflected in the different legal status of torture, on the one hand, and cruel, inhuman, or degrading treatment, on the other. The British government, recognizing the significance of this distinction, appealed a ruling by the European Court and scored a public relations victory by reducing the court’s finding to the judgment that Britain was “only” guilty of cruel and unusual treatment of Irish political prisoners—not actual torture (Millett, 1994, p. 102). The CAT also creates a distinction, requiring states parties to criminalize official torture, but simply exhorting them to work to prevent cruel, inhuman, or degrading treatment.

Yet, a 2007 study of survivors of torture found that psychological manipulation, humiliating treatment, exposure to adverse environmental conditions, and forced stress positions were no different from physical torture in terms of the level of traumatic stress and the long-term psychological after-effects that they cause. Based on their findings, the authors conclude that it is misleading to distinguish between torture and other cruel, inhuman, or degrading treatment since both can cause severe mental suffering (Başoğlu et al., 2007, pp. 283–284). Other medical experts concur that psychological manipulations and forced stress positions are not “substantially different from physical torture in terms of the severity of mental suffering” and can have severe, long lasting effects (McCoy, 2011, p. 32; Wolfendale, 2009, pp. 50–51). Similarly, in a study of the effects of extreme stressors on the brain, neuropsychologist Shane O’Mara (2015, p. 4) finds that coercive practices that involve extreme and prolonged stress impair the normal functioning of the brain and result in neuropsychiatric disorders, disrupting if not permanently damaging bodily function and thereby violating even the standards laid down by administration lawyers. Along the same lines, David Luban (2009, p. 222) argues that “there is something deeply wrong, not to mention perverse, about the entire enterprise of trying to draw fine lines between torture and lesser abuses. An essential continuity exists between them, because all have the degradation of their victim as their core.” For Luban, the key characteristic of torture lies in the abject humiliation of the victim, whose sense of complete powerlessness creates such intense fear that the victim is terrorized and “broken.” For interrogators like Tony Lagouranis and Eric Fair, interrogation practices such as stress positions and sleep deprivation, despite their designation as legal rules of engagement, actually did constitute torture; the practices stripped the detainees of their humanity, with effects that were intense, severe and extraordinarily painful (Fair, 2016; Lagouranis, 2008, pp. 335–336).

Elaine Scarry shares Luban’s view that torture renders the victim completely powerless, but she focuses on the way that the infliction of excruciating pain does this. According to Scarry (1985, pp. 27–28), torture reduces people to prisoners of their bodies, demonstrating and magnifying the power of the agent (and, by extension, the regime) employing torture. The “annihilating power of pain” narrows victims’ consciousness, reducing their focus to immediate bodily sensations. Their bodies become agents of their own agony, as the “grotesque overload” of physical pain disintegrates their sense of self and eliminates their ability to express and project themselves through language (Scarry, 1985, pp. 47–49). The “world-destroying” effect of torture, in turn, compels prisoners to confess, “to assent to words that through the thick agony of the body can be only dimly heard” (Scarry, 1985, pp. 33, 35). The confession, as Scarry explains, is what provides a justification for brutal treatment, redirecting moral responsibility from the torturer to the prisoner. Even though the torturer is in complete control, the interrogation process shifts blame on to the victim, pretending that the prisoner has caused this situation by withholding information. Yet even a confession does not absolve the prisoner: “despite the fact that in reality he has been deprived of all control over, and therefore all responsibility for, his world, his words, and his body, he is to understand his confession as it will be understood by others, as an act of self-betrayal” (Scarry, 1985, p. 47).

Along similar lines, Jean Améry, a victim of torture during World War II, writes that torture destroys people’s sense of self by breaking down the boundaries of the body. When the victim’s body is attacked—and furthermore, when the victim has neither the ability to defend himself nor the expectation of help from another person—then he becomes nothing but a body in pain: “only in torture does the transformation of the person into flesh become complete. Frail in the face of violence, yelling out in pain, awaiting no help, capable of no resistance, the tortured person is only a body, and nothing else beside that” (Améry, 1980, p. 28). The victim’s body becomes the instrument of the torturer, who exercises absolute domination over the prisoner’s flesh and spirit, rendering the victim helpless and hopeless (Améry, 1980, pp. 21–40).

Recent philosophical essays argue that it is the asymmetrical relationship between victims and torturers that makes torture distinctive. Whereas victims of torture are completely vulnerable and exposed, their torturers are in perfect control. Victims are utterly at the mercy of their tormentors; the prisoner “cannot effectively evade, retaliate or shield himself against these assaults. Unlike other kinds of attack, here the victim must simply take it: there is no reply or counter open to him” (Sussman, 2005, p. 31). Moreover, the victim cannot stop his or her physical suffering by admitting defeat; although in theory torture will stop once the subject confesses or provides certain information, “a torturer seldom if ever knows how much useful information the tortured has or how much the tortured must confess in order to have confessed ‘everything’” (Davis, 2005, p. 164). Torture takes full advantage of victims’ helplessness, turning them into active accomplices in their own debasement. Because of the insistent, primal demand for relief from intense pain, victims cannot help but search for some way of appeasing or mollifying their tormentors; the “victim experiences within himself a dialectic where some part of him serves as the eager agent of his tormentor . . . the victim finds in his pain, and his own immediate responses to that pain, a surrogate for the torturer. The victim’s own voice, the voice of his body, has come in part to speak the torturer’s mind” (Sussman, 2005, p. 24). In addition to exploiting the victim’s pain, torturers defile, degrade, and overwhelm their victims with shame, similar to victims of rape; not surprisingly, sexual torture is often used to emphasize the power of the tormentor and the vulnerability of the victim (Millett, 1994, pp. 34–35).

The essence of torture is the destruction of the victim’s sense of agency. The deliberate infliction of severe physical pain or mental suffering is not enough; this must be done with the purpose of bringing individuals to such a point of weakness that they betray themselves. The victim must be broken, forced to offer information against his or her will (Bellaby, 2015, p. 187). When a person voluntarily submits to painful procedures (such as self-flagellation or painful medical procedures), this does not qualify as torture. Moreover, torture differs from coercion insofar as torture seeks to terrorize victims into submission by overwhelming their capacity to exercise rational control over their decisions. Finally, torture differs from corporal punishment insofar as the latter prescribes a specific, predetermined penalty for a particular transgression and does not seek to break the guilty party’s will (Miller, 2017; for a contrasting view, see Schabas, 1996, p. 4). As Manfred Nowak (2006, p. 832), the UN Special Rapporteur on Torture, explains in his study of US and international standards of torture, what distinguishes torture is the total subordination of the victim to the will and power of the torturer, for example through prolonged incommunicado detention in a secret place, which permits the perpetrator to intentionally inflict pain or suffering so as to extract a confession, obtain information, or punish or intimidate the victim.

History of Torture

Although torture has come to be widely rejected, this was not always the case. Historically, torture was not only common in times of war and social upheaval, but it was also openly practiced in many societies as an integral part of the judicial system. No euphemisms for torture were needed in medieval Europe since there was no need to deny the process and torture enjoyed general cultural consent (Silverman, 2001, p. 21). Throughout much of history and much of the world, torture was seen as an effective technique for obtaining “true” information as well as an appropriate punishment for the immoral and a useful deterrent against future misconduct. Ancient Greeks accepted torture for the interrogation of slaves based on the assumption that slaves could not be trusted to reveal the truth voluntarily. The Romans adopted this practice and extended it to citizens, including Christians who were “put to the question” to force them to renounce their faith. With the rise of Christianity, torture fell into relative disuse until the 11th century, when European judicial systems resurrected the practice of judicial torture and the Catholic Church reversed its previous opposition to torture and sanctioned the use of torture against heretics (Peters, 1985, pp. 13–14). In China, torture was a legally sanctioned means of extracting information and confessions from the Han dynasty (206 bce–220 ce) until just before the end of the imperial system in 1905. Chinese officials were convinced that “the measured use of torture could result in more substantive justice—in the sense that the guilty were more likely to be convicted and the innocent allowed to go free” (Park, 2008, p. 37).

Analyzing the accepted use of torture in medieval Europe, Lisa Silverman (2001, p. 61) explains that Europeans generally believed that the truth could be elicited from the subconscious by applying physical pain: “It was widely believed that the body had many ways to betray the criminal involuntarily, speaking the truth in signs for all to see while the will stopped the tongue. Pallor, for example, was well known to indicate guilty knowledge.” By inflicting physical pain, legalized torture was therefore seen to bypass the human will and force truth from the guilty, while God would reward those proven innocent. Contrary to the modern assumption that only testimony that is given voluntarily is true, early modern people assumed that “the accused spoke the truth not freely but under compulsion, and it was precisely this compulsion … which made evidence acquired under torture so valuable” (Silverman, 2001, p. 66).

The legal acceptability of torture therefore rested on the belief that it was an unpleasant but necessary means for discovering the truth and thus achieving justice. Criminals would be made to confess their guilt and pay for their crimes; the innocent would be vindicated. In both cases, the pain and suffering of those tortured would help achieve a higher goal, whether preserving civilization, fulfilling a sacred religious mission, or protecting the community. Torture was also used as an accepted part of ordinary criminal procedure in Europe from the thirteenth to the late eighteenth centuries, when judges were required to establish certain guilt in order to convict someone of a serious crime; certainty, in turn, was established through the testimony of two eyewitnesses or the accused person’s own confession. Because the two-eyewitness standard was so difficult to achieve, torture became an accepted means of extracting confessions (Langbein, 2004, pp. 94–97).

In each case, the practice of torture was subject to strict rules, including the specific instruments to be used, the way in which they were to be applied, and the types of people who could be subjected to torture. In each case, initial restrictions were eventually eased. In ancient Greece and Rome, the rule that only slaves could be tortured was eventually expanded to include other groups; similarly, initial exemptions for women, children, and privileged classes were eventually dropped in medieval Europe. As John Conroy (2000, pp. 27–28) explains, “the class of people whom society accepts as torturable has a tendency to expand.”

Changes in European legal systems led to a general ban on torture in continental Europe over the course of the 18th century. With the advent of less severe punishments for crimes, standards of proof could be relaxed. Defendants could be sentenced to jail, the workhouse, or exile rather than death; whereas the high stakes involved in sentencing a prisoner to death required definitive proof, relatively lighter sentences could be justified on the basis of circumstantial evidence. Torture-induced confessions were no longer legally justified (Langbein, 2004, pp. 97–99). The abolition of a legalized system of judicial torture also reflected a shift in thinking about torture. Previously, torture was justified on religious grounds since it was presumed to benefit the sufferer by forcing him to admit the truth, even against his will, and thus bringing him closer to God. The infliction of pain was seen as a positive technique for saving the souls of fallen Christians, allowing sufferers to atone for their sins and win eternal salvation; hence, torture in the Inquisition was justified as Rettungsfolter or salvation-oriented torture (Brunkhorst, 2009, p. 75; Glucklich, 2001).

Over the course of the 18th century, however, Enlightenment thinkers challenged this sacramental view of pain, contending that pain had no redeeming value and no connection to metaphysical truths; these intellectuals replaced the sacramental vision of pain with a medical approach to pain that emphasized the need to relieve suffering. Rather than embracing physical pain as a positive technique for overcoming selfishness, they condemned it as a negative practice that destroyed the self. They denied the value of testimony elicited through torture and charged that torture was a “tool of despotism” and “a weapon in the arsenal of political oppression” (Silverman, 2001, p. 171). In his 1764 “Essay on Crimes and Punishments,” Italian prison reformer Cesare Beccaria wrote that governments have no right to authorize the punishment of a citizen so long as there remains any doubt of his guilt; according to Beccaria, torture was “a sure way to acquit robust scoundrels and to condemn weak but innocent people” (quoted in Foot, 2006, p. 135). The changing cultural landscape meant that, in Europe at least, torture was transformed from a generally accepted practice to a generally rejected practice by the end of the 18th century. In fact, judicial torture had already become less common before this time, since judges had previously become skeptical that torture necessarily produced truthful testimony (Langbein, 2004, p. 99; Silverman, 2001, pp. 66–67).

Michel Foucault analyzes a similar shift in attitudes toward the use of torture as a method of punishment. Prior to the 18th century, torture was not only used to extract confessions but was also used as a form of extreme punishment intended to demonstrate and strengthen the sovereign’s power: “Its aim is not so much to re-establish a balance as to bring into play, at its extreme point, the dissymmetry between the subject who has dared to violate the law and the all-powerful sovereign who displays his strength” (Foucault, 1975, pp. 48–49). As such, torture was used to terrorize the rest of the population by demonstrating the fearsome consequences of incurring the sovereign’s wrath. Yet the sovereign’s public use of torture as punishment, for example in public executions, did not necessarily deter others from misbehaving; it sometimes triggered sympathy for the convict and prompted riots in support of the prisoner. Public torture therefore proved counterproductive to the goals of the state, undermining rather than reinforcing the sovereign’s power. As such, it gave way to new disciplinary techniques that rejected brutal violence as an ineffective method of controlling and manipulating behavior. Although there was still a fear of violence and brutality, prisoners were found to be much more effectively disciplined by isolating them and subjecting them to constant surveillance.

Changes in legal theory meant that after the 18th century, judicial torture was no longer condoned in Western Europe. Coerced confessions were considered unreliable and were inadmissible as evidence. Torture came to be seen as the hallmark of a premodern, uncivilized culture, and modern, liberal regimes were assumed to have abandoned such a barbaric practice. According to the author of a 1907 entry on torture in the Encyclopaedia Britannica, “the whole subject is now only one of historical interest as far as Europe is concerned” (quoted in Twining & Twining, 1973, p. 305). When Rejali (quoted in Parry, 2005, pp. 517–518) wrote that the use of torture in Iran was not an anachronism but part of a rational, bureaucratic structure typical of modern states, his association of torture and modernity prompted one critical reviewer to ask: “if the growth of torture in 20th-century Iran and its changing forms are caused by efforts to modernize, why do we not torture in the modern United States or Western Europe?”

In fact, Rejali documents the continued use of torture by modern democracies in his 2007 book, Torture and Democracy, and argues that the scrutiny of civil society actors and human rights advocates has not ended torture but has incentivized interrogators to employ non-scarring techniques instead: “When we watch interrogators, interrogators get sneaky” (Rejali, 2007, p. 9). In his comparative and historical analysis of torture, Christopher Einolf (2007) provides statistical evidence that there was a resurgence of torture in the 20th century. Torture increased with the rise of fascist and communist regimes in Germany, the Soviet Union, and their allies. These regimes employed torture as a mechanism for social control, using it to sow fear in their populations and neutralize dissent. Changes in the intensity and nature of military conduct in the first half of the century also led to an increase of torture of prisoners of war and occupied populations: “When total war tactics were combined with ideological and nationalist disrespect for conventional limitations on war, massacre, violence against civilians, and torture of enemy civilians and prisoners of war occurred at unprecedented levels” (Einolf, 2007, p. 114).

Yet torture was also a common social practice used against domestic populations, especially members of marginalized groups (Parry, 2005, p. 521). In the United States, a 1931 report by the National Commission on Law Observance and Enforcement documented the widespread use of stress positions, battery, psychological torture and the “water cure” in order to obtain information about a crime, highlighting the systematic use of the “third degree” by police and prison officials (Coyne & Hall, 2018, pp. 146–147). Some of these techniques were apparently first tested in the Philippines where a wide range of torture techniques became standard operating procedure, subject to few constraints and no legal accountability (Coyne & Hall, 2018, pp. 142–145).

Despite international efforts to promote human rights at the conclusion of World War II, torture continued to happen “off stage” in colonial and foreign locations and against members of marginalized groups whose guilt was presumed and credibility was questioned (Parry, 2005, pp. 521–522). While colonial governments practiced torture prior to the 20th century, the rise of anti-colonial movements in the second half of the 20th century prompted colonial authorities to tighten their control and resort to increasingly brutal methods, including torture. Counter-insurgency campaigns against guerrilla forces also led to an increase in torture, as insurgents and their supporters were more likely to be tortured for information than prisoners of war in previous conventional wars, who were believed to possess relatively little information of value to the other side (Einolf, 2007, p. 114). The 20th century also saw the rise of military governments in Latin America and communist governments in Asia that used systematic repression and torture to crack down on real and suspected opponents. These governments used torture and state terror to subdue their populations, exaggerating the extent and depth of security threats due to rigid ideological frameworks that convinced leaders that repression of the general population was necessary (Pion-Berlin & Lopez, 1991, p. 68).

Since torture is forbidden under international law, it is commonly associated with nondemocratic regimes that fail to recognize limits on their power and wantonly subject individuals to brutal, inhumane treatment. However, with the rise of international campaigns against torture and the conditioning of foreign aid on adherence to human rights norms, nondemocratic states came to appreciate the value of appearing to conform to international norms prohibiting torture. Even nondemocratic states that regularly employ torture have ratified the UN CAT, calculating that doing so will enhance their reputations without actually preventing them from continuing to use torture (Hathaway, 2004, pp. 202–208). In addition, both nondemocratic and democratic states have turned to “clean” torture techniques in order to avoid bad publicity about their human rights practices (Rejali, 2007, pp. 23–26). These techniques, which leave no visible marks and are therefore easier to downplay and deny, were first developed in democratic countries, where governments continued to believe that coercive interrogation techniques were a useful means of generating valuable intelligence but also realized that they were being watched and judged by others in how well they respected human rights. Thus, in order to avoid bad publicity and preserve a veneer of legitimacy, they turned to what Rejali (2007, pp. 1–10, 2011, p. 30) refers to as stealth torture: “whenever there’s a free press, church groups, and politicians to watch the police, the interrogators literally pull their punches” and use clean torture techniques instead.

Thus, democratic and undemocratic countries alike have developed ways of evading public scrutiny of unsavory practices that they officially renounced but privately employed. Historian Alfred McCoy (2011, p. 31) examines the program of psychological torture developed and propagated by the CIA during the Cold War, emphasizing that the techniques of sensory disorientation and self-inflicted pain proved advantageous insofar as they did not leave “clear signs of abuse, greatly complicating any investigation, prosecution, or attempt at prohibition.” The CIA conducted research on mind control through “truth drugs” like LSD and experimented with forms of psychological torture through extreme sensory deprivation in hopes of gaining an advantage vis-à-vis their communist rivals (Coyne & Hall, 2018, pp. 155–160). In Vietnam, where constraints were absent or lax, torture was used as part of Project Phoenix, a program designed to find, capture, interrogate and kill Vietnamese sympathetic to the Viet Cong. Some of the torture techniques developed in Vietnam were later brought back to the United States and used to coerce confessions from criminal suspects, as in the case of Vietnam veteran and Chicago police officer Jon Burge, reported to have overseen the torture of hundreds of African American suspects (Coyne & Hall, 2018, pp. 162–166). Other cases of democratic countries’ use of torture against people defined as insurgents, terrorists, revolutionaries, or criminals include the French use of torture in Vietnam and Algeria, British practices in Kenya and Northern Ireland, dirty war tactics employed by Spanish security forces against Basque separatists, as well as methods used by Israeli security services to interrogate Palestinians suspected of “hostile terrorist activity” (Conroy, 2000, pp. 4–8; Millett, 1994, pp. 74–116; Parry, 2005, p. 518; Supreme Court of Israel, 1999). Furthermore, officials from intelligence agencies of democratic countries were also involved in training their counterparts in developing countries on interrogation methods, including torture; the US Army School of the Americas (SOA) trained Latin American militaries to use practices such as torture, extortion, kidnapping and execution (Blakeley, 2006; McCoy, 2011, p. 33; Quigley, 2011, pp. 54–57).

These practices were deliberately hidden from the public since this allowed democratic citizens to imagine that the methods employed by their government were both efficient and moral; as Andrew Linklater (quoted in Steele, 2010, p. 153) explains, “concealment protects moral sensibilities.” Democratic elites also took advantage of the public’s greater willingness to condone the use of torture against racial minorities, immigrants and other marginalized citizens, who are often seen as “deserving” rough treatment since they are members of groups that are portrayed as suspicious or even dangerous (Kelly, 2009; Wolfendale, 2009, p. 58). As Brent Steele (2010, p. 153) explains, if we do not identify—or identify with—the individuals who are subjected to torture, then we can imagine that they are as bad as possible and that they deserve the treatment they receive. Einolf (2007, p. 102) agrees, citing evidence that torture is used “more often against people who are not full members of a society, such as slaves, foreigners, prisoners of war, and members of racial, ethnic, and religious outsider groups” or when the state is perceived to be under severe threat. As a result, despite domestic and international condemnation, torture techniques continue to be used not only in cases of national emergency but also as a routine part of many judicial systems (Tse, 2011).

Political and Institutional Influences on the Practice of Torture

All types of governments respond to violent challenges with repression. Studies have found that governments are significantly more likely to resort to torture when they face threats to their continued rule, especially when they are engaged in civil and international wars, face a violent dissident challenge, or are attacked by a transnational terrorist group (see, for example, Conrad et al., 2017; Davenport et al., 2007; Einolf, 2007; Wantchekon & Healy, 1999). Erwin Staub (2003) studied social indicators of genocide and torture, analyzing Turkey during years of Armenian genocide (1914–1918), Nazi Germany (1933–1945), Pol Pot’s Cambodia (1976–1979), Argentina’s “Dirty War” (1976–1983) and identified the scapegoating of a subgroup as the social condition most conducive to rise of torture. Criminologist Ronald Crelinsten (2005, pp. 76–77) adds that torture is more likely in times of national emergency when there is an especially strong sense of threat to security and when prevailing ideology touts a “sacred mission” that provides a justification for violating standard social norms; he finds that torture is also more common when large numbers of suspects need to be processed. James Piazza and James Walsh (2009, pp. 126–145) come to a more nuanced conclusion in their analysis of the impact of terrorism on government protections of human rights, finding that countries experiencing a large number of terrorist attacks were more likely to engage in disappearances and extrajudicial killings but not in torture, political imprisonment or restrictions on free speech and assembly. They do admit, however, that states may respond to terrorist attacks by engaging in more “clean” torture techniques that are by definition difficult for human rights monitoring groups to detect, much like disappearances and extrajudicial killings (Piazza & Walsh, 2009, p. 138).

According to Emilia Powell and Jeffrey Staton (2009, pp. 149–150), 83% of all states that ratified the CAT engaged in at least minimal treaty violations, while 42% of ratifiers—including 30% of democracies that ratified the CAT—systematically violated the convention. Democracies are therefore not as exceptional as they often like to think of themselves. On the one hand, liberal democratic institutions have been found to limit state coercion and repression. Democracies have a better record when it comes to respecting physical integrity rights (including freedom from torture as well as lack of extrajudicial killings, disappearances, and political imprisonment), especially where they have effective constitutional guarantees of the right to fair and public trials (Keith, Tate, & Poe, 2009, p. 652; Simmons, 2009, pp. 273–274). In democracies, a free media can act as a watchdog and expose human rights violations (Whitten-Woodring, 2009), contested elections can incentivize public officials to protect physical integrity rights (Cingranelli & Filippov, 2010) and effective judicial institutions can punish instances of state torture (Powell & Staton, 2009). Nonetheless, democratic governments have not abandoned the use of torture and many consider it useful in case of external threats such as transnational terrorism. In such instances, they turn to stealth torture and clean torture techniques in order to maintain plausible deniability and reduce the risk of exposure (Rejali, 2007). This allows democratic officials to claim that they remain within the bounds of legally acceptable levels of violence, as in former CIA director Michael Hayden’s central metaphor in his 2016 memoir Playing to the Edge, which he describes as playing so close to the line that you get chalk dust on your cleats. In fact, as Jinee Lokaneeta (2011, pp. 34–35) argues, jurisprudence in liberal democracies uses such ambiguity to unequivocally reject torture while simultaneously permitting other forms of excess that exist at the border of legality and illegality. Thus, while constitutional provisions and human rights treaty commitments are not mere “parchment barriers” to human rights abuses including torture (Keith et al., 2009), they do not prevent states from violating such commitments in the face of terrorism or violent dissent.

Elections and liberal democratic institutions do, however, tend to reduce the duration of human rights abuses once violent challenges end. Courtenay Conrad and Will Moore (2010) find that elections and liberal democratic institutions can influence governments to stop using torture once the threat of violent opposition is gone; states with popular suffrage and a free press are generally more likely to terminate their use of torture because allegations of torture are more likely to become public and the executive is more likely to be held responsible and removed from office. Under such circumstances, executives have an incentive to take actions to convince interrogators and jailers that they will be caught and punished if they resort to torture. Democratic systems with a greater number of checks on executive authority (that is, a large number of veto players) are less likely to change torture practices, however, since the system is weighted toward maintaining the status quo. For example, though Barack Obama promised to close the detention camps at Guantanamo Bay, he was blocked from doing so by Congress, showing that institutionalized separation of powers makes it difficult to change human rights policy.

Thus, the use of torture may continue even when the threat of violent dissent is removed: “States whose agents engage in torture in a given year have a 93% chance of continuing to torture in the following year” (Conrad & Moore, 2010, p. 459). The use of torture is often embedded in local culture and can be difficult to dislodge, given that executives cannot monitor and control all of the agents assigned to interrogate and supervise prisoners. In interviews with local police in India, Rachel Wahl (2014, pp. 820–821) found that a deeply entrenched moral beliefs that defended the use of torture against hardened criminals and terrorists who were seen as “less human.” These enforcement officers rejected the premise that their job was to uphold the law and protect human rights; instead, they justified the use of torture as necessary to fight evil and achieve justice (Wahl, 2014, pp. 821–822). In the face of genuine moral beliefs that conflict with human rights norms, upholding the personal integrity rights of criminals and marginalized individuals can be difficult even under normal circumstances. However, governments and enforcement officers who refrain from using torture are likely to maintain good practices “until a national security threat justifies the rougher handling of detainees” (Simmons, 2009, pp. 305–306).

Democratic institutions may constrain governments from engaging in human rights abuses but they may also respond to popular pressure to use harsh tactics against foreigners and marginalized groups, rewarding leaders who vow to protect national security. Democratic majorities expect their leaders to protect them and elected officials who approve or condone the use of torture in the name of order and national security are often re-elected: “blocks of voters who explicitly support aggressive coercive behavior against unwanted others are able to elect executives who promise to take off the gloves on the grounds that they are implementing the will of the people” (Conrad, Hill, & Moore, 2018, p. 4). Using data from Amnesty International on specific allegations of torture and ill treatment, Courtenay Conrad, Daniel Hill and Will Moore (2018, p. 13) find that contested elections do not lead to greater protection of individual human rights; popularly “elected leaders are expected to protect the majority, and elections may incentivize them to permit/order state agents do so by committing human rights violations against people perceived to be threatening.” Even where there is widespread support for the general prohibition of torture, voters are unlikely to make protection of human rights a key electoral issue, especially for the rights of people of different racial, ethnic and cultural backgrounds. Survey research suggests, for example, that Americans are more accepting of government torture when a detainee has an Arabic name and when the alleged crime is terrorism (Conrad, Croco, Gomez, and Moore, 2018; Piazza, 2015). Contrarily, powerful constitutional courts are created to protect the rights of political minorities by monitoring and punishing violations of constitutional and human rights. Where courts are powerful, victims may file cases against human rights violators and those found guilty may be held accountable. Even so, effective courts cannot entirely prevent repressive practices, as officials may maintain plausible deniability by employing clean torture techniques that are easier to hide: “Courts may decrease state repression writ large, but they also appear to encourage executives and their agents to be more clever in hiding human rights violations that can be narrowly targeted by turning to clean torture when they do occur” (Conrad et al., 2018, pp. 13–14). Similarly, despite expectations that constitutional prohibitions on torture should counter majoritarian support for repression of minorities, Adam Chilton and Mila Versteeg (2015) find that countries with torture bans actually engaged in more torture than countries without such bans and that torture levels actually worsened in the ten years after bans were adopted.

Just as liberal democracies generally have better human rights records but nonetheless practice torture in response to violent threats such as transnational terrorism, their compliance with international human rights treaties varies. Here, too, independent domestic courts may not be enough to deter governments from violating their commitment to refrain from using torture, and majoritarian pressures may actually reward some governments in the form of increased popular support for governments that openly violate international law (Lupu & Wallace, 2019). Yonatan Lupu (2013) argues that even powerful courts in liberal democracies may lack sufficient information to secure prosecutions for alleged cases of torture where evidence is difficult to obtain and standards of proof are high; as a result, ratification of the ICCPR has improved government respect for the freedoms of speech, association, assembly, and religion, but has not reduced the extent to which governments use torture, extrajudicial killings, political imprisonment, and disappearances.

Nondemocratic governments, however, are more likely to flout the terms of international human rights treaties, even when they have ratified such treaties. Such governments may cynically ratify treaties as a form of “social camouflage” to avoid international criticism even though they have little intention of actually improving their human rights practices (Simmons, 2009, p. 112). Nondemocratic governments are actually more likely to practice torture if they have signed the CAT than if they have not (Hathaway, 2004). James Vreeland (2008, pp. 69–70) explains this initially counterintuitive finding by analyzing the different circumstances facing “open” and “closed” dictatorships. He found that dictatorships with the trappings of competitive party politics are both more likely to sign the CAT and also demonstrate higher rates of torture. He explains that “closed” dictatorships in which power is concentrated in a single political party, junta, or leader actually face less opposition since defection is invariably punished. Such regimes face little pressure to ratify human rights treaties; moreover, since they rely on fear and intimidation to rule, “even a symbolic gesture against torture could introduce ambiguity over [their] limitations” (Vreeland, 2008, p. 78). Dictatorships that allow some competition, however, face much greater prospects of defection and consequently have higher average rates of torture. At the same time, the greater level of uncertainty that characterizes multiparty dictatorships means that domestic political actors are in a better position to press their government to ratify human rights conventions such as the CAT. Vreeland (2008, p. 94) clarifies that becoming a state party to the CAT does not itself cause an increase in torture and “may even serve to help reduce torture as governments adopt CAT provisions into domestic law”—which is why the domestic opposition in multiparty authoritarian regimes push for accession in the first place. Jay Goodliffe and Darren Hawkins (2006, pp. 359–360) also argue that ratification entails a significant commitment insofar as it requires ratifying states to verify policy implementation, and establishes international monitoring and delegates prosecuting authority to other states through universal jurisdiction.

Emilie Hafner-Burton (2008, pp. 700–701) argues that “naming and shaming” by human rights NGOs, media sources, and international organizations is often followed by even more acts of torture and disappearances. This arguably results from the fact that international pressure compels repressive governments to make a variety of improvements in political rights, but doing so exacerbates leaders’ insecurity and prompts them to engage in increased terror to offset these improvements, especially when armed opposition groups or elections threaten their hold on power (Hafner-Burton, 2008, p. 712). While Beth Simmons (2009, p. 305) acknowledges that international human rights treaties like the CAT cannot force governments to comply, she insists that these treaties can and do reduce torture in polities that have at least moderate levels of public accountability:

CAT ratification resonates in those polities; individuals and groups who may have good reason to fear mistreatment of themselves, their families, their countrymen, or other humans by the government have strong incentives to mobilize to implement the international ban in domestic law.

Overall, respect for constitutional protections of personal integrity rights and for the global norm against torture depends to a large extent on the expectation of norm enforcement and accountability, which is greater but not given in democratic states. Strong domestic legal systems can place significant constraints on the behavior of government leaders and agents, but enforcement is particularly difficult in the case of torture since violations tend to be hidden and public outcry is sporadic at best.

Conclusion

Although disagreements remain over the precise definition of torture, its modern use, and the effectiveness of international treaties prohibiting torture, most scholars agree that “the impressive worldwide prohibition of torture must be honored rather than co-opted by legal sleight of hand” (Miller, 2005, p. 41). Though torture continues to be practiced by modern states, including democracies, it rejects the liberal tenet of the inviolable dignity of the individual and gives agents of the state absolute power over prisoners. This has a corrupting influence and is incompatible with liberalism and limited government. Even if the intended end is to save lives, the use of torture is generally rejected as an unacceptable means, and government resort to hypocrisy, secrecy and obfuscation, insisting that their enhanced interrogation methods do not qualify as torture. The absolute prohibition against torture under domestic and international law incentivizes officials to turn to clean techniques that have the veneer of acceptability even though they can cause excruciating pain and severe mental suffering.

Continued use of such techniques is made easier by downplaying their seriousness and restricting their use to marginalized groups in remote locations. “When officials declare that they have not approved illegal or inappropriate treatment, their denials tend to be accepted because we are likely to overlook or discount the seriousness of the treatment that has been permitted or inflicted” (Parry, 2005, p. 521). The use of clean torture means that victims have a difficult time proving that they were brutally mistreated and they “may be already defined as suspect or untrustworthy because of their race or ethnicity, or because they were detained in the first place” (Parry, 2005, p. 521). Moreover, torture is easier to hide when victims are others—foreign enemies or racial, ethnic, religious, and other minorities who are naturally regarded with less sympathy and more suspicion. People are less likely to define an act as torture when it is perpetrated against others and are therefore less likely to be concerned about torture when it occurs overseas against terrorism suspects that are viewed as a distinct “other” or is perpetrated by a member of one’s out-group (Kearns, 2015, p. 10; Norris, Larsen, & Stastny, 2010).

Distinguishing between in-groups and out-groups is also one of the steps involved in conditioning individuals to commit violence and torture. In-group solidarity is fostered through common uniforms, rules, and language. Camaraderie and peer pressure urge individuals on while authorities encourage obedience and provide rewards (Huggins, Haritos-Fatouros, & Zimbardo, 2002). Socialization and group cohesiveness increase anonymity and decrease self-awareness, fostering a lack of personal accountability for abusive behavior (Costanzo & Gerrity, 2009, pp. 196–198). In his study of a German police unit during World War II, Christopher Browning (1998) argues that the “ordinary men” who served in the unit were gradually desensitized to brutality by the examples of their peers and the mockery they feared if they failed to participate. Hostility toward out-groups is encouraged by dehumanizing victims and characterizing them as a threat, making it less disturbing to use violence against them. As Janice Gibson and Mika Haritos-Fatouros (1986) analyze in their study of the brutal but effective process of desensitizing special police units in Greece to torture suspected communists, military training gradually desensitized soldiers to violence and reduced the strain normally created by repugnant acts. The moral disengagement that made violence possible involved giving the enemy derogatory names and portraying them as less than human, making it easier to kil1 them (Gibson & Haritos-Fatouros, 1986, p. 55). Similarly, in the Stanford prison experiment, the students who were randomly assigned to be guards quickly became aggressive and abusive within the prison, insulting and bullying the prisoners:

Even without training, the student guards “knew” from television and movies that they were supposed to punish prisoners; they “knew” they were supposed to feel superior; and they “knew” they were supposed to blame their victims. Their own behavior and that of their peers gradually numbed their sensitivity to what they were doing, and they were rewarded by the power they had over their prisoners. (Gibson & Haritos-Fatouros, 1986, p. 57)

Dehumanization also helps explain the lack of accountability even when cases of torture are uncovered. In John Conroy’s (2000) study of the “unspeakable acts” committed by British soldiers against Northern Irish men falsely accused of being Irish Republican Army terrorists, he notes that the British government took years to acknowledge that the abuse and continued to deny that the techniques constituted torture. Moreover, the government charged without evidence that the tortured detainees were nothing more than “thugs and murderers” while “the perpetrators, instigators, and defenders of the five techniques . . . escaped unscathed, their reputations untarnished” (Conroy, 2000, p. 187).

The continued use of torture and abusive practices also relies on social amnesia, which perpetuates a culture of impunity (Rajali, 2011, p. 27). Occasional scandals tend to quickly fade from memory and torture retreats back to the shadows, at the margins of prevailing political discourse (Parry, 2005, p. 521). For example, the British public’s knowledge of torture and other abuses committed by British colonial authorities in Kenya has clearly faded:

It is fair to say that even now, fifty years later, the British public is not really aware of what went on . . . British ignorance about Mau Mau is of a peculiar, resilient kind. It is breached every so often, but then heals over again. . . . There was a period in the later 1950s when everyone knew, or could know, what was going on. . . . All that seems to have been forgotten. The British need to believe that their Empire was run and eventually dismantled with restraint and humanity—as opposed to the disgusting brutality of the French, Dutch, Belgian, Portuguese, Spanish, and German colonial empires. (Neal Ascherson quoted in Parry, 2005, p. 521)

Just as it is wrong to see torture as an historical anachronism, it is also misleading to claim that it is a new temptation or the product of exceptional circumstances (Athey, 2007). Historical amnesia and general lack of concern for the possible torture of others make it difficult to hold governments accountable for violations and pressure them to comply with the prohibition against torture. When no one is held responsible, torture is able to continue; as Rejali (2011, p. 38) notes, “nothing predicts future torture as much as past impunity.” Despite international agreements and constitutional prohibitions, governments and their citizens have not demanded effective action to halt the spread of torture; in fact, they have been able to pretend that torture only occurs on an exceptional basis by relegating torture to hidden places and condoning its use against threatening and marginalized groups. This protects officials who ordered and implemented torture from responsibility. In order to confront torture, people must concern themselves with rights of all others, even those from “the torturable classes” (Conroy, 2000, p. 251).